Original scientific paper
https://doi.org/10.22572/mi.30.2.1
Assessing Sentiments Towards Artificial Intelligence in the Croatian Media Using the ChatGPT Artificial Intelligence Tool
Ivana Erceg Matijašević
; PhD Student of Media and Communication, University North
Martina Baričević Debelec
; PhD Student of Media and Communication, University North
Ljerka Luić
orcid.org/0000-0002-2707-9367
; University North
Abstract
Algorithms and artificial intelligence (AI) tools are crucial for digital literacy and competitiveness in today’s high-tech environment, transforming jobs across various sectors, including the media. The history of journalism is closely linked with technological development; from Gutenberg’s printing press to radio, television, the internet and AI tools. This paper aims to determine the perception of AI in the Croatian media and its alignment with Croatian journalists’ opinions, as well as assess ChatGPT’s effectiveness in sentiment analysis. The insights gained should improve the understanding of the impact of AI on public opinion and its ethical implications for journalism. The research methodology combined AI-driven sentiment analysis (ChatGPT Plus) with qualitative content analysis. A total of 45 articles about AI published in the Croatian media between April and September 2023 were evaluated. Two hypotheses were tested: H1, that ChatGPT’s sentiment analysis matches human assessment, and H2, that the Croatian media generally express a positive view of AI. The results confirmed H1, with ChatGPT’s sentiment analysis corresponding to human assessment in 44 out of 45 cases. H2 was partially corroborated; sentiment distribution pointed to neutral (42%), positive (36%), and negative (22%) views of AI. Positive articles highlighted the benefits of AI in areas such as healthcare, while negative articles raised privacy and employment concerns. The results point to a balanced perspective on AI in the Croatian media, recognising advantages and risks alike. Future research can expand on these findings by examining long term sentiment trends in domestic media and comparing them with global trends. Improved AI tools for independent data collection and more accurate sentiment interpretation can further support this endeavour. Additionally, this paper paves the way for further sentiment analysis of media articles, exploring the implications of AI application on readers’ perception and reception of information.
Keywords
artificial intelligence; ChatGPT; digital competitiveness; digital intelligence; journalism; sentiment
Hrčak ID:
325154
URI
Publication date:
19.12.2024.
Visits: 0 *